INTERSTELLAR SPACE ABBOT 215 



diffuse nebulae. He was able to derive the probable density of the 

 interstellar matter from considerations of the constitution of the 

 diffuse nebulae. Without following the steps of Eddington's analy- 

 sis, we may summarize by saying that he arrived at the conclusion 

 that as far as the velocity of their particles was concerned these 

 nebulae would behave as if raised to a temperature of 10,000° Abs.C. 

 The resulting density of the interstellar gases, he found, is of the 

 order of 10"-* g/cm^. This density, as Eddington points out, must 

 be a maximum value, as otherwise the velocities of the stars, owing 

 to the increased mass of the stellar system, would be greater than 

 observed. 



The extreme rarity of interstellar diffuse matter, as Eddington 

 showed, was favorable to the high speeds necessary for the splitting 

 off of one or more electrons from calcium atoms when illuminated by 

 the rays of the stars so that they would be in the state to show the 

 H and K lines in their spectra. To explain his point fully would 

 take us into the study of photoelectricity, which would be too far 

 afield. I will therefore only ask the reader to accept the fact that it 

 is the frequency of vibration, not the intensity of rays of light, which 

 determine whether or not they can singly or multiply ionize atoms. 

 The rays of the hottest stars, however distant and thereby enfeebled, 

 possess the tremendous frequency of vibration required. 



The property of absorbing, nearly totally, specially selected rays 

 from stars is adapted to give the ionized calcium atoms velocities of 

 interaction far above those of molecules of a solid body situated in 

 interstellar space. Such specially influenced absorbing atoms radi- 

 ate and absorb as if at very high temperatures. Computed by the 

 ordinary laws of the perfect radiator or " absolutely black body ", 

 space has a theoretical temperature of only 3° C. above absolute zero. 

 Calcium atoms at 3° Abs.C. would not be ionized and would not 

 show the H and K violet spectral lines. It is the ionizing influence 

 of high-frequency rays from stars that puts these atoms in condition 

 to absorb H and K rays. 



The crucial test of Eddington's hypothesis is that the intensity of 

 the stationary lines must be a simple function of the distance. 



The intensity of a stationary calcium line may be measured with 

 a microphotometer. Comparing the contour thus derived with the 

 contour produced by a known amount of calcium vapor in the 

 laboratory, we can determine the total number of ionized calcium 

 atoms between the observer and the star. For a star about 10,000 

 or 1.5,000 light-years away we find lO^'^ ionized calcium atoms in a 

 column having a section of 1 square centimeter. Remembering that 

 1 out of 1,000 calcium atoms is in the singly ionized state competent 

 to produce the observed H and K lines, we conclude that the total 



